Kurtosis is another way we can talk about what a probability distribution looks like on a chart, with particular emphasis on its tails and peaks. Distributions are said to have excess kurtosis when kurtosis is greater than three because a univariate normal distribution will have kurtosis of exactly three. A distribution that does not exhibit any excess kurtosis is called mesokurtotic:
Leptokurtotic – taller and thinner peak as compared to a normal mesokurtotic distribution. Positive excess kurtosis.
Platykurtotic – flatter, wider peak than a normal distribution. Negative excess kurtosis.